1 

CHAPTER C31 WIND TUNNEL PROCEDURE

2 

3  Wind tunnel testing is specified when a building or other structure contains any of the
4  characteristics defined in Sections 27.1.3, 28.1.3, 29.1.3, or 30.1.3, or when the designer wishes to
5  more accurately determine the wind loads. For some building or structure shapes, wind tunnel
6  testing can reduce the conservatism caused by the enveloping of wind loads inherent in the
7  Directional Procedure, Envelope Procedure, or Analytical Procedure for Components and

LY
8  Cladding (C&C). Also, wind tunnel testing accounts for shielding or channeling and can more
9  accurately determine wind loads for a complex building or structure shape than the Directional

N
10  Procedure, Envelope Procedure, or Analytical Procedure for C&C can. It is the intent of the

O
11  standard that any building or other structure can be allowed to use the wind tunnel testing method
12  to determine wind loads.
n
io
13  Requirements for proper wind tunnel testing are given in ASCE 49. Such a standard is a
14  prerequisite for determining wind loads, either via numerical or physical wind tunnel testing.
at

15  While Computational Fluid Dynamics (CFD) simulations are increasingly being used in Wind
m

16  Engineering applications, ASCE 49 does not explicitly identify all procedures necessary for CFD.
or

17  While awaiting a similar standard documenting the procedures needed to obtain reliable and
18  accurate wind loads using CFD tools, any use of CFD to determine design main wind force
nf

19  resisting system (MWFRS), C&C or other structures’ wind loads, requires peer review and a
rI

20  verification and validation (V&V) study (Yeo 2020). In the absence of a standard, this is necessary
21  to address quality assurance and quality control of this method.
Fo

22  When using CFD as a numerical wind tunnel, many of the requirements described in ASCE 49 for
23  physical wind tunnel testing are applicable. For example, a suitable approach flow, accurate
24  geometry, the inclusion of significant nearby structures, and consideration of the potential for
25  modal excitation and aeroelastic effects are all also needed in the numerical model. Once validated
26  against a base-case physical model, the CFD simulation can help resolve details that cannot be
27  measured in the physical model, and/or allow for sensitivity analysis for parametric changes. In
28  the absence of such validation, the numerical wind tunnel simulations can only be considered as
29  qualitative information.
1 
 
 1  It is common practice to resort to wind tunnel tests when design data are required for the following
2  wind-induced loads:

3  1. Curtain wall pressures resulting from irregular geometry;

4  2. Across-wind and/or torsional loads;
5  3. Periodic loads caused by vortex shedding; and
6  4. Loads resulting from instabilities, such as flutter or galloping.
7  Boundary-layer wind tunnels, capable of developing flows that meet the conditions stipulated in
8  Section 31.2, typically have test-section dimensions in the following ranges: width of 6 to 12 ft (2

LY
9  to 4 m), height of 6 to 10 ft (2 to 3 m), and length of 50 to 100 ft (15 to 30 m). Maximum wind
10  speeds are ordinarily in the range of 25 to 100 mi / h (10 to 45 m / s ) The wind tunnel may be

N
11  either an open-circuit or closed-circuit type.

O
12  Three basic types of wind tunnel test models are commonly used. These are designated as follows:
13 
n
(1) rigid pressure model (PM), (2) rigid high-frequency base balance model (H-FBBM), and (3)
io
14  aeroelastic model (AM). One or more of the models may be used to obtain design loads for a
15  particular building or structure. The PM provides local peak pressures for design of elements, such
at

16  as cladding and mean pressures, for the determination of overall mean loads. The H-FBBM
m

17  measures overall fluctuating loads (aerodynamic admittance), for the determination of dynamic
responses. When motion of a building or structure influences the wind loading, the AM is used for
or

18 
19  direct measurement of overall loads, deflections, and accelerations. Each of these models, together
nf

20  with a model of the surroundings (proximity model), can provide information other than wind
rI

21  loads, such as snow loads on complex roofs, wind data to evaluate environmental impact on
22  pedestrians, and concentrations of air pollutant emissions for environmental impact
Fo

23  determinations. Several references provide detailed information and guidance for the
24  determination of wind loads and other types of design data by wind tunnel tests (Cermak 1977,
25  Reinhold 1982, ASCE 1999, Boggs and Peterka 1989).

26  Wind tunnel tests frequently measure wind loads that are significantly lower than those required
27  by Chapters 26, 27, 28, 29, and 30 because of the shape of the building or other structure; the
28  likelihood that the highest wind speeds occur at directions where the building or structure’s shape
29  or pressure coefficients are less than their maximum values; specific buildings or structures
30  included in a detailed proximity model that may provide shielding in excess of that implied by
2 
 
 1  exposure categories; and necessary conservatism in enveloping load coefficients in Chapters 28
2  and 30. In some cases, adjacent buildings or structures may shield the subject building or structure
3  sufficiently so that removal of one or two of the adjacent buildings or structures could significantly
4  increase wind loads. Additional wind tunnel testing without specific nearby buildings or structures
5  (or with additional buildings or structures if they might cause increased loads through channeling
6  or buffeting) is an effective method for determining the influence of adjacent buildings or
7  structures.

8  For this reason, the standard limits the reduction that can be accepted from wind tunnel tests to

LY
9  80% of the result obtained from Part 1 of Chapter 27, Part 1 of Chapter 28, or Chapter 30, if the
10  wind tunnel proximity model included any specific influential buildings or other objects that, in

N
11  the judgment of an experienced wind engineer, are likely to have substantially influenced the

O
12  results beyond those characteristic of the general surroundings. If there are any such buildings or
13  objects, supplemental testing can be performed to quantify their effect on the original results and
14  n
possibly justify a limit lower than 80%, by removing them from the detailed proximity model and
io
15  replacing them with characteristic ground roughness consistent with the adjacent roughness. A
at
16  specific influential building or object is one within the detailed proximity model that protrudes
17  well above its surroundings, is unusually close to the subject building, or may otherwise cause
m

18  substantial sheltering effect or magnification of the wind loads. When these supplemental test
or

19  results are included with the original results, the acceptable results are then considered to be the
20  higher of both conditions.
nf

However, the absolute minimum reduction permitted is 65% of the baseline result for C&C and
rI

21 
22  50% for the MWFRS. A higher reduction is permitted for MWFRS because C&C loads are more
Fo

23  subject to changes caused by local channeling effects when surroundings change, and they can
24  easily be dramatically increased when a new adjacent building is constructed. It is also recognized
25  that cladding failures are much more common than failures of the MWFRS. In addition, in the case
26  of MWFRS, it is easily demonstrated that the overall drag coefficient for certain common building
27  shapes, such as circular cylinders (especially with rounded or domed tops), is one-half or less of
28  the drag coefficient for the rectangular prisms that form the basis of Chapters 27, 28, and 30.

29  For C&C, the 80% limit is defined by the interior Zones 1 and 4 in Figures 30.3-1, 30.3-2A–C,
30  30.3-3, 30.3-4, 30.3-5A–B, 30.3-6, 30.3-7, and 30.4-1. This limitation recognizes that pressures in

3 
 
 1  the edge zones are the ones most likely to be reduced by the specific geometry of real buildings,
2  compared with the rectangular prismatic buildings assumed in Chapter 30. Therefore, pressures in
3  edge and corner zones are permitted to be as low as 80% of the interior pressures in Chapter 30
4  without the supplemental tests. The 80% limit based on Zone 1 is directly applicable to all roof
5  areas, and the 80% limit based on Zone 4 is directly applicable to all wall areas.

6  The limitation on MWFRS loads is more complex because at any point, the load effects (e.g.,
7  member stresses or forces, deflections) are the combined effect of a vector of applied loads instead
8  of a simple scalar value. In general, the ratio of forces, moments or torques (force eccentricity), at

LY
9  various floors throughout the building using a wind tunnel study will not be the same as those
10  ratios determined from Chapter 27 and 28, and therefore comparison between the two methods is

N
11  not well defined. Requiring each load effect from a wind tunnel test to be no less than 80% of the

O
12  same effect resulting from Chapters 27 and 28 is impractical and unnecessarily complex and
13  detailed, given the approximate nature of the 80% value. Instead, the intent of the limitation is
14  n
effectively implemented by applying it only to a simple index that characterizes the overall
io
15  loading. For flexible (tall) buildings, the most descriptive index of overall loading is the base
at
16  overturning moment. For other buildings, the overturning moment can be a poor characterization
17  of the overall loading, and the base shear is recommended instead.
m
or

18  C31.4 LOAD EFFECTS FOR SITE-SPECIFIC BUILDINGS, OTHER STRUCTURES,

19  AND COMPONENTS
nf

20  C31.4.1 Mean Recurrence Intervals of Load Effects

rI

Examples of analysis methods for combining directional wind tunnel data with the directional
Fo

21 
22  meteorological data, or probabilistic models based thereon, are described in Lepage and Irwin
23  (1985), Rigato et al. (2001), Isyumov et al. (2013), Irwin et al. (2005), Simiu and Filliben (2005),
24  and Simiu and Miyata (2006).

25  C31.4.2 Limitations on Wind Speeds

4 
 
 1  Section 31.4.2 specifies that the statistical methods used to analyze historical wind speed and
2  direction data for wind tunnel studies shall be subject to the same limitations specified in Section
3  31.4.2 that apply to the Analytical Method.

4  Database-Assisted Design Wind tunnel aerodynamics databases that contain records of pressures
5  measured synchronously at large numbers of locations on the exterior surface of building models,
6  have been developed by wind researchers, such as Simiu et al. (2003) and Main and Fritz (2006).
7  Such databases include data that permit a designer to determine, without specific wind tunnel tests,
8  wind-induced forces and moments in MWFRSs and C&C of selected shapes and sizes of buildings.

LY
9  A public domain set of such databases, recorded in tests conducted at the University of Western
10  Ontario (Ho et al. 2005, St. Pierre et al. 2005) for buildings with gable roofs, is available on the

N
11  National Institute of Standards and Technology website, www.nist.gov/wind (NIST 2012).

O
12  Interpolation software for buildings with similar shape, and dimensions close to and intermediate
13  between those included in the set of databases, is also available on that website. Because the
14  n
database results are for generic surroundings, as permitted in ASCE 49, interpolation or
io
15  extrapolation from these databases should only be used if Condition 2 of Section 27.1.2 is true.
at
16  Extrapolations from available building shapes and sizes are not permitted, and in some instances,
17  interpolations may not be advisable. For these reasons, the guidance of an engineer who is
m

18  experienced in wind loads on buildings and is familiar with the usage of these databases is
or

19  recommended.
nf

20  All databases must have been obtained using testing methodology that meets the requirements for
wind tunnel testing, as specified in Chapter 31.
rI

21 
Fo

22  C31.4.3 Wind Directionality.

23  The variability of wind speed determined for particular azimuth intervals is greater than that of the
24  wind speed determined regardless of wind direction (Isyumov et al. 2013). Consequently, wind
25  loads and wind-induced effects, determined by allowing for wind directionality, are inherently less
26  certain. Several methods for combining data from wind tunnel model studies with information on
27  wind speed and direction at the project site are currently in use (Isyumov et al. 2013, Yeo and
28  Simiu 2011, Simiu 2011). Whichever method is used shall be clearly described to allow scrutiny
29  by the designer and the Authority Having Jurisdiction. A common approach for allowing for

5 
 
 1  uncertainties in the wind direction is to rotate the project wind climate relative to the orientation
2  of the building or structure. This rotation of the wind climate at the building location is intended
3  to ensure that the wind loads determined for design are not unconservative and shall be considered,
4  regardless of the method used for arriving at the design wind speeds. The appropriate magnitude
5  of wind climate rotation varies depending on the quality and resolution of the directional wind
6  climate data at the project site.

7  C31.4.5 Limitations on Wind Loads for Ground-Mounted Fixed-Tilt Solar Panel Systems

LY
8  The minimum C&C and MWFRS wind loads, as indicated in ASCE 7, are primarily applicable to
9  buildings and are not applicable to ground-mounted fixed-tilt solar panel systems. The limitations

N
10  contained herein are to establish the lower bound wind loads for wind tunnel studies of conditions

O
11  similar to those addressed in Section 29.4.5. The limits on wind tunnel results shown in Figure
12  29.4-10 represent an envelope of the static wind pressures measured in the wind tunnel. The limits
13 
n
on wind tunnel results shown in Figure 29.4-11 represent an envelope of the dynamic wind loads
io
14  derived from the wind tunnel pressure data, which necessarily included simplifying (conservative)
at
15  assumptions regarding the supporting structure and dynamic properties of the system. Specific
16  installations or system geometries/supporting structures may give significantly lower loads than
m

17  those in Figures 29.4-10 and 29.4-11; limits are imposed to prevent too much deviation from the
or

18  enveloped results. Ground-mounted fixed-tilt solar PV systems can have wind-tunnel-based wind
19  loads less than the lower bound thresholds indicated in Sections 31.4.5. In order to use these lower
nf

20  values, a peer review of the test and report is required.

rI

21  C31.5 LOAD EFFECTS FOR BUILDINGS, OTHER STRUCTURES, AND

Fo

22  COMPONENTS USED AT MULTIPLE SITES

23  C31.5.1 Wind Loads

24  For ASCE 7-22, the wind tunnel test requirements were renamed to account for wind tunnel tests
25  on generic buildings, other structures, and components that are used in multiple locations or on
26  multiple buildings. Roof-mounted solar arrays are but one such example. Other examples include
27  building-mounted components, such as sunshades, HVAC units, screen walls, or could be free-
28  standing, such as ground-mounted solar trackers, gazebos and fences.

6 
 
 1  In determining wind loads on generic buildings, other structures, and components, the approach
2  needs to be similar to that used to develop the ( GC p ) figures in ASCE 7 by modeling the generic

3  buildings with various features to capture a wide range of effects. The objective of such testing is
4  to evaluate aerodynamic effects accounted for by pressure coefficients (in contrast to site-specific
5  wind tunnel testing, which also evaluates the effect of surrounding structures and terrain). Nearby
6  buildings should not be included, unless they are to be a part of every design application for such
7  buildings, other structures, or components.

8  Wind tunnel testing must include a sufficiently large test matrix to address an appropriate range of

LY
9  the relevant variables that affect wind loads, as listed in the provisions. Guidance for testing is
10  provided in ASCE 49. Wind loads are expressed as coefficients usable in Chapters 27, 29, and 30

N
11  to produce loads in engineering units. Alternately, a different formulation of nondimensional load

O
12  coefficients may be used, provided that the analysis procedure is clearly defined in the test report.

13  n
C31.5.2 Limitations on Wind Loads for Rooftop Solar Collectors. 
io
14  In ASCE 7-22, the scope of wind tunnel testing for buildings, other structures, and components
at

15  that are used at multiple sites was added. A peer review is required for the use of this approach,
m

16  except for items that are already covered in the standard, such as roof-mounted solar collectors.
However, peer review is required when the loads fall below the minimum threshold.
or

17 

Regarding rooftop solar collectors, the minimum components and cladding wind load pressures
nf

18 
19  indicated in ASCE 7 are primarily applicable to the building envelope and are not applicable to
rI

20  rooftop solar collectors. The limitations contained herein are to establish the lower bound wind
pressures for wind tunnel studies of conditions similar to those addressed in Figure 29. 4-7. The
Fo

21 
22  limits on wind tunnel results shown in Figure 29.4-7 represent an envelope of wind loads measured
23  in the wind tunnel without deflectors or shrouds that are commonly used to lower wind loads.
24  Specific installations or collector geometries may give significantly lower loads than Figure 29.4-
25  7; limits are imposed to prevent too much deviation from the enveloped results. Solar collector
26  systems that have aerodynamic devices or more efficient profiles can have wind-tunnel-based wind
27  loads less than the lower bound thresholds indicated in Sections 31.5.2 and 31.5.3. In order to use
28  these lower values, a peer review of the test and report is required.

7 
 
 1  C31.5.3 Peer Review Requirements for Wind Tunnel Tests of Buildings, Other Structures,
2  and Components at Multiple Sites 

3  This section provides the requirements for peer reviews of wind tunnel studies. The peer reviewer’s
4  qualifications and requirements are included to promote consistencies among the various
5  jurisdictions so that a peer review could be accepted by multiple enforcement agencies. The peer
6  reviewer’s qualifications are intended to be those of a wind tunnel expert who is familiar with wind
7  tunnel testing of buildings and the applicability of the ASCE 7 provisions to determine generalized
8  wind design coefficients for roof-mounted solar collectors. One source for peer reviewers is the

LY
9  American Association for Wind Engineering’s (AAWE) boundary layer wind tunnels list
10  (http://www.aawe.org/info/wind_tunnels.php).

N
O
11 

12  REFERENCES
n
io
13  ASCE. 1999. “Wind tunnel model studies of buildings and structures.” ASCE Manuals and
at
14  Reports of Engineering Practice No. 67, Reston, VA: ASCE.
m

15  ASCE. 2012. “Wind tunnel testing for buildings and other structures.” ASCE/SEI 49-12, Reston,
VA: ASCE
or

16 

17  Boggs, D. W., and Peterka, J. A. 1989. “Aerodynamic model tests of tall buildings.” J. Eng. Mech.
nf

18  115(3), 618–635.

rI

19  Cermak, J. E. 1977. “Wind-tunnel testing of structures.” J. Eng. Mech. Div. 103(6), 1125–1140.
Fo

20  Ho, T. C. E., Surry, D., Morrish, D., and Kopp, G. A. 2005. “The UWO contribution to the NIST
21  aerodynamic database for wind loads on low buildings: Part 1. Archiving format and basic
22  aerodynamic data.” J. Wind Eng. Indust. Aerodyn., 93, 1–30.

23  Irwin, P., Garber, J., and Ho, E. 2005. “Integration of wind tunnel data with full scale wind
24  climate.” In Proc., 10th Americas Conf. on Wind Eng., Baton Rouge, LA. doi:
25  10.1061/541X.0000654.

8 
 
 1  Isyumov, N., Ho, E., and Case, P. 2013. “Influence of Wind Directionality on Wind Loads and
2  Responses.” In Proc., 12th Americas Conf. on Wind Eng. 141(8) doi: 10.1061/541X.0001180,
3  04014208.

4  Isyumov, N., Mikitiuk, M., Case, P., Lythe, G., and Welburn, A. 2013. “Predictions of wind loads
5  and responses from simulated tropical storm passages.” Proc., 11th Int. Conf. on Wind Eng, D. A.
6  Smith and C. W. Letchford, eds. 19(3), 295–320. doi: 10.12989/was.2014.19.3.295.

7  Kopp, G., and Banks, D. 2013. “Use of the wind tunnel test method for obtaining design wind

LY
8  loads on roof-mounted solar arrays.” J. Struct. Eng. 139(2), 284–287.

9  Kopp, G. A., Farquhar, S., and Morrison, M. J. 2012. “Aerodynamic mechanisms for wind loads

N
10  on tilted, roof-mounted, solar arrays.” J. Wind Eng. Ind. Aerodyn . 111, 40–52.

O
11  Kopp, G., Maffei, J., and Tilley, C. 2011. “Rooftop solar arrays and wind loading: A primer on
using wind tunnel testing as a basis for code compliant design per ASCE 7.” London, ON:
12 
n
io
13  Boundary Layer Wind Tunnel Laboratory, University of Western Ontario, Faculty of Engineering.
at
14  Lepage, M. F., and Irwin, P. A. 1985. “A technique for combining historical wind data with wind
15  tunnel tests to predict extreme wind loads.” In Proc., 5th US Nat. Conf. on Wind Eng., M. Mehta,
m

16  ed. doi: 10.1061/541X.0001625, 04016148.

or

17  Main, J. A., and Fritz, W. P. 2006. “Database-assisted design for wind: Concepts, software, and
nf

18  examples for rigid and flexible buildings.” NIST Building Science Series 180. Washington, DC:
19  National Institute of Standards and Technology.
rI

20  NIST (National Institute of Standards and Technology ). 2012. “Extreme winds and wind effects
Fo

21  on structures.” Accessed March 5, 2012. www.nist.gov/wind.

22  Reinhold, T. A., ed. 1982. “Wind tunnel modeling for civil engineering applications.” In Proc.,
23  Int. Workshop on Wind Tunnel Modeling Criteria and Techniques in Civil Eng. Applications,
24  Gaithersburg, Maryland. New York: Cambridge University Press.

25  Rigato, A., Chang, P., and Simiu, E. 2001. “Database-assisted design, standardization, and wind
26  direction effects.” J. Struct. Eng., 127(8), 855–860.

9 
 
 1  Simiu, E. 2011. Design of building for wind. Hoboken, NJ: Wiley.,

2  Simiu, E., and Filliben, J. J. 2005. “Wind tunnel testing and the sector-by-sector approach to wind
3  directionality effects.” J. Struct. Eng . 131(7), 1143–1145.

4  Simiu, E., and Miyata, T. 2006. Design of buildings and bridges for wind: A practical guide for
5  ASCE Standard 7 users and designers of special structures. Hoboken, NJ: Wiley,.

6  Simiu, E., Sadek, F., Whalen, T. A., Jang, S., Lu, L.-W., Diniz, S. M. C., et al. 2003. “Achieving
7  safer and more economical buildings through database-assisted, reliability-based design for wind.”

LY
8  J. Wind Eng. Indust. Aerodyn., 91, 1587–1611.

St. Pierre, L. M., Kopp, G. A., Surry, D., and Ho, T. C. E. 2005. “The UWO contribution to the

N
9 
10  NIST aerodynamic database for wind loads on low buildings: Part 2. Comparison of data with

O
11  wind load provisions.” J. Wind Eng. Indust. Aerodyn., 93, 31–59.

12  n
Yeo, D. 2020. “A summary of industrial verification, validation, and uncertainty quantification
io
13  procedures in computational fluid dynamics. NIST Internal Report 8298. Gaithersburg, MD:
at
14  National Institute of Standards and Technology.
m

15  Yeo, D., and Simiu, E. 2011 “High-rise reinforced concrete structures: Database-assisted design
16  for wind,” J. Struct. Eng. 127, 1340–1349.
or
nf
rI
Fo

10